World’s first sub-1B parameter model with functional tool calling capability. Generates structured JSON execution plans for tool/plugin orchestration. Given a user request and available tools, it produces a valid JSON object specifying which tools to call, with what parameters, and in what order. This model was trained using a novel multi-stage approach developed by AIREV: — Base model: Qwen 3.5-0.8B — 47K+ curated samples with reasoning traces and structured JSON outputs — Real production plugin schemas from the OnDemand platform (2,176 plugins) — LLM-evaluated data quality filtering (score >= 8/10) — Full fine-tune (not LoRA) — research shows full FT outperforms at sub-1B scale — Group Relative Policy Optimization with a novel 4-phase progressive reward curriculum —
Модальности:
Генерация текста
Области применения:
Диалог / чат Вызов функций (Tool use)
Задача: Генерация текста
Автор: airev-ae
Теги: qwen3_5_text, tool-calling, function-calling, agent, edge-ai, mobile, GRPO, agentic
Лайков: 4 | Загрузок: 162
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
